Students who learn to decode with at least a minimal ability are faced with an equally daunting task in learning to understand what they have read. Teachers often assume students will develop these skills naturally. Although there are effective programs that directly teach important comprehension skills, teachers do not have the necessary training or time to implement these programs. For the student, failure to learn reading comprehension skills commonly creates long term social, political, and economic problems. This project will incorporate advanced speech recognition technology with the computer delivery of an effective, research-proven reading comprehension program to effectively and economically teach basic comprehension skills to remedial and at-risk students at the middle school, high school, and community college levels. In Phase I, the first ten lessons of the program will be produced and evaluated with low-performing students using a speech recognizer trained by our staff for local dialect. In Phase II, we will develop the complete 100-lesson speech recognition remedial reading comprehension program. The program will include a speech recognition engine, complete multimedia presentation of instructional content, printed materials and all necessary hardware. PROPOSED COMMERCIAL APPLICATION: This project will lead to the Phase II production of a 100-lesson remedial reading comprehension program for students in middle school, high school, and adult literacy programs. The product will be attractive to schools and learning centers that serve low?literacy students, as well as to the home market and industry training centers.